82 
Thais, however, did not lose the girth entirely and so become a sus¬ 
pended pupa like the Nymphalids. It is, as we have seen, one of several 
forms, that do not lead directly to Parnassius in lineal order, but appear 
to have branched off from the Papilionids at the same time. In the 
Pierid we have a Papilionid that has lost the power of lateral move¬ 
ment, probably in connection with the method of suspension by a girth. 
The Nymphalid is a Pierid that has lost the girth. This close connec¬ 
tion of the Pierids and Nymphalids is one that has not apparently been 
hitherto insisted on. The form of the egg is the same in both, and 
differs from that of the Papilionids, which retains the smooth dome¬ 
shaped form characteristic of the egg of the Hesperids. Again, the 
restriction of movement to the lateral direction is very remarkable, for 
it occurs nowhere else; and, since it exists throughout both these 
nearly related families, it seems reasonable to assume that it is a common 
inheritance and not a separate acquisition, especially as it is obviously 
useful to the Pierid but of no obvious value to the Nymphalid, in 
which it persists only because movement once lost is never regained 
(at least this rule holds so generally true, that we may reasonably apply 
it here in explanation.) But, accepting this explanation, it follows, that 
the divergence must have taken place whilst the Pierids still retained 
both the 5th and 6th abdominal as movable segments, as is still the 
case in Aporia, Delias (figs. 6-7-8) etc.; and further, seeing that the 
Nymphalid preserves the double nose-horn of the Papilionid it must 
have taken place before the Pierid was reduced to a single median one, 
and it is interesting to note that Delias, one of the earliest Pierids, still 
affords some evidence of the former presence of a double nose-horn 
in the Pierids. Our own A. cratcegi does so in a less degree. 
A consideration of these pupal characteristics, as well as of those of 
the eggs, leads me to believe that Pierid and Nymphalid started 
together from the Papilionid, shortly afterwards separating, and thence¬ 
forward pursuing very parallel courses. It would follow that, in spite 
of the method of suspension of the pupa and of sundry imaginal 
characters, the Pierids are nearly if not quite as much entitled as the 
Nymphalids to rank as a separate family from the Papilionids, and to 
be associated with the Nymphalids rather than with the Papilionids. 
As we pass from the Aporinae to the higher sub-families of Pieridae, we 
find progressive loss of movement. In the Pierinae only the fifth seg¬ 
ment remains movable ; the same condition obtains in the Bhodocerinae, 
and in the Anthocarinae the pupa is solid. I have not obtained an y 
Pierid pupa with only one incision moveable. Whilst I would separate 
Aporia from Pieris as distinctly as I would Bhodocera from Anthocaris 
(Euchloe ), it seems incorrect to associate Pieris with Bh odocera because both 
happen to have the same formula of segmental mobility. The curved 
form of pupa, due to the ventral bulging of the wings and the shortness 
of the antennae, associates Bhodocera and Anthocaris, but places them 
apart from the others. By analogy with other families, I should expect 
to find the Anthocarine pupa separated from the Pierine, whilst both 
were still at the Aporine stage, that is with both 5th and 6tli abdominal 
segments movable, and that the genealogical tree would not be thus— 
1, Aporia ; 2, Pieris ; 3, Bhodocera ; 4, Anthocaris, but rather as under, 
the blanks representing forms that probably exist though unknown to 
me. 
